On some relationships between correlation analysis and similar-fade analysis results of drift measurements in the lower ionosphere

Abstract

In the comparisons hitherto made between similar-fade analysis results and correlation analysis results of ionospheric drift measurements two essential facts seem to have been left out of account, namely 1. (i) that there are at least two different ways of making similar-fade analysis which often yield quite different magnitudes of drift velocity, and 2. (ii) that the similar-fade method is necessarily confined preferably to the short-period components of the fading and that, therefore, its results may not be a priori comparable with the results of a correlation analysis which includes the long-period fading components as well. From investigations of ionospheric drift measurements in the LF range at medium latitude it can be shown that there are close relationships between similar-fade and correlation analysis results, if correlation analysis is performed after submitting the original fading curves to a mathematical filter which suppresses the long-period fading components. The drift velocity obtained from one variant of the similar-fade method (the method of averaging time delays) is very similar to the term “apparent drift velocity” in the correlation analysis, whereas the drift velocity obtained from the other variant of the similar-fade method (the method of averaging drift vectors) is very similar to the term “true drift velocity” in the correlation analysis. Even the term “chaotic velocity” can approximately be found from the similar-fade method by combining the results of the two variants according to a formula borrowed from the correlation analysis formalism. These relationships, though not yet founded mathematically, are expected to be generally valid also at higher frequencies and in other latitudes, as far as there is no extreme anisotropy of the ionospheric irregularities.